ES2955369A1 - IMAGE RECORDING, LOCATION AND RECOVERY SYSTEM FOR VIDEO SURVEILLANCE AND METHOD FOR SUCH SYSTEM (Machine-translation by Google Translate, not legally binding) - Google Patents

Abstract

System (1) for recording, locating and retrieving images for video surveillance, which comprises: - a plurality of video surveillance cameras (2), - particular storage units (2a) in a loop of the videos taken, associated with each of them. the cameras (2), -a motion detection module (2b) associated with each of the cameras (2), -an indexing server (3), provided with a memory (3a) for storing representative images (100 , 101, 102) extracted from the videos taken by each camera (2), provided with a controller (4) for displaying the images, associated with a display screen (5), - a management interface (4a), and -communications links (6) that are connecting each camera (2) and its particular storage unit (2a) with the indexing server (3). Furthermore, it comprises a method for said system that works with a limited number of images when searching. (Machine-translation by Google Translate, not legally binding)

Description

DESCRIPTION

Recording system, location and recovery of images for video surveillance and method for said system

Object of the Invention

The present invention refers to an image recording and recovery system for video surveillance, and to a specific method for using said system that allows greater speed and efficiency in the communications network with respect to the volume of data transferred.

Background of the invention

Currently, continuous recording and downloading of the video signal is carried out in video surveillance systems, and reviews are carried out manually by technical personnel on all of the videos taken, with the possibility of increasing the speed and time jump of viewing the video. recording, or through event detections.

However, high activity in the monitored area (for example, a street with high traffic of vehicles or people) gives rise to a high number of events, normally referred to as motion detection, and also does not discriminate their type, so there is a massive amount of information, among which you have to look for what is likely to be the object of the search.

In some cases, software can analyze the videos stored on the centralized server - where all the videos are downloaded - and help identify passages where events have occurred (that is, where motion is detected), but this does not alleviate the requirements of video storage capacity on the centralized server, nor does it greatly simplify the manual and personal work, which in most cases must view multiple video passages that are of no interest.

Description of the invention

The image recording and recovery system for video surveillance of the invention comprises:

- a plurality of video surveillance cameras (understood as such in this document to be cameras that take videos of any public or private space that you wish to monitor, which may be of interest to identify incidents, events or similar; - particular units ( which may or may not be incorporated into the same camera) loop storage of the videos taken, associated with each of the cameras,

- a motion detection module associated with each of the cameras,

- an indexing server, provided with a storage memory for representative images extracted from the videos taken by each camera and sent to said indexing server, to facilitate the location of events of interest, provided with an image display controller ,

- a management interface (local or rather a station or terminal connected to the indexing server), and associated with a display screen of the representative images stored on the server with their associated date and time, for image retrieval or additional videos associated with the representative images from the cameras' particular storage units, and

- communications links (wired or wireless) that are connecting each camera and its particular storage unit with the indexing server, for the transmission of representative images to the server, and for receiving the related videos stored in the particular storage units. storage.

As representative images, in this document it is understood that they are a set of images or frames, each of them sent to the indexing server every time period set in the system, when movement is detected in the motion detection module of the camera. concrete, in such a way that if no movement is detected, no representative images are sent, and if movement is detected, only these representative images are sent to the server, which show specific moments each time period adjusted in the system, of the scene taken by the camera.

The system of the invention works with the method of recording and recovering images for video surveillance of the invention, which comprises:

- a stage of registration and indexing of representative images, which includes the following substages:

- loop video recording (i.e. taking video fragments of a given duration and continuously replacing them with the next video fragment at the end of a given period, such that storage capacity is only required for a video that It is always updated in a certain period of time (for example 30 days), from a plurality of cameras, which are covering spaces intended to be monitored, and storage of said videos in particular storage units of each camera,

- detection of movements (scene variations) in the monitored spaces,

- in case of positive movement detection, send representative images, extracted from the videos taken each adjusted sending time period, to an indexing server, through communications links that are connecting each camera and its unit particular storage with said indexing server, and - storage of the representative images received on the indexing server, in a storage memory;

and comprising an image search and display stage that comprises the following substages:

- introduction of a date (and time), or a range of dates and times, of retrieval of representative images stored on the indexing server, through the management interface of a display controller for the images of the indexing server,

- search, on the indexing server, for the representative images close to the date and time entered or the range of dates and times entered, through the image display controller, and presentation of said representative images retrieved on a display screen display, - manual selection, through the display controller, between the recovered representative images with an initial temporal distance between them of, at at least, a marker image (image that shows a fact of interest regarding the event or events sought), and

- downloading, from the particular storage unit of each camera associated with the selected marker image, a video fragment containing said marker image.

In this way, the server's storage capacity requirements are reduced, leaving the videos stored distributedly in the cameras; and not only that, it also allows working with a limited number of images when performing manual searches, being able to index the representative images with other representative images closer in time to one where an event of interest is seen, to refine the search by iteration, and to finally recover the video of interest from the particular storage unit of the camera, which results in great speed by working mainly with images instead of videos, resulting in reduced traffic of communications in the entire process of locating and downloading said video fragments.

Brief Description of the Drawings

Figure 1.- Shows a view of the system of the invention

Figure 2.- Schematically shows the iteration stages in the search for the representative marker image.

Description of the Preferred Embodiment

The system (1) for recording and retrieving images for video surveillance of the invention comprises (see Fig. 1):

- a plurality of video surveillance cameras (2), which respectively include particular loop storage units (2a) for the videos taken, associated with each of the cameras (2), an associated motion detection module (2b). to each of the cameras (2).

- an indexing server (3), provided with a memory (3a) for storing representative images (100, 101, 102) extracted from the videos taken by each camera (2), and also provided with a controller (4) display of images,

- a management interface (4a), for searching among the representative images (100, 101, 102) stored on the server (3) * and for retrieving videos associated with the representative images (100, 101, 102) from the units particular storage (2a) of the cameras (2), with the corresponding display screen (5), and

- communications links (6) that are connecting each camera (2) - and particular storage unit (2a) - with the indexing server (3), for the transmission of representative images (100, 101, 102) to the server and for receiving the related videos stored in the particular storage units (2a).

The motion detection module (2b) may comprise a comparator of successive images, to detect movements due to variations in the images taken, or It may directly comprise a motion detector associated with one or more cameras, to detect movements in the area covered by each camera.

For its part, the method of recording and retrieving images for video surveillance of the invention comprises:

- a stage of registration and indexing of representative images, which includes the following substages:

- loop video recording (i.e. taking video fragments of a given duration and continuously replacing them with the next video fragment at the end of a given period, such that storage capacity is only required for a video that is always updated in the period of time - for example 30 days - determined), from a plurality of cameras (2), of spaces intended to be monitored, and storing said videos in particular storage units (2a) of each camera ( 2).

- detection of movements in the monitored spaces,

- in case of positive movement detection, send representative images (100, 101, 102) extracted from the videos taken each adjusted sending time period, to an indexing server (3), through links (6 ) of communications that are connecting each camera (2) and its particular storage unit (2a) with said indexing server (3), and

- storage of the representative images (100, 101, 102) received in the indexing server (3) in a storage memory (3a),

- and an image search and display stage that includes the following substages:

- introduction of a date and time or a range of dates and times of retrieval of representative images stored on the indexing server (3), through a controller (4) for displaying the images of the indexing server (3), - search on the indexing server (3) for representative images (100, 101, 102) close to the date and time entered, or to the range of dates and times entered, through the controller (4) for displaying the images. images, and presentation of said first representative images (100, 101, 102) recovered on a display screen (5),

- manual selection, through the display controller (4), between the representative images (100, 101, 102) recovered from at least one marker image (A), and

- download, from the particular storage unit (2a) of each camera (2) associated with the selected marker image (A), of a video fragment (A') that contains said marker image (A).

Preferably, the procedure will comprise that, the search for first representative images (100), which will ideally be the closest ones prior to the date (which also includes the time) entered, or prior to each date of a range of dates entered, as well as the presentation of said first representative images (100) recovered on the display screen (5). That is, the moment before the possible event is sought.

The search for the first representative images (100) may comprise the recovery of the closest image prior to the entered date and a series of first representative images (100) following this one at intervals of a determined initial time, and their presentation on the display screen (5). In this way, starting from a specific moment, a limited series of still images can be displayed that range from a moment immediately prior to the event until a time has elapsed that includes said event.

To refine the search, prior to the sub-step of selecting the marker image (A) (or representative marker images), the procedure may comprise one or more sub-steps of iteration of searching for successive representative images (101, 102), increasingly closer. in time to a previous representative image (see Fig. 2). That is, if in the first recovery, the first representative images (100) were spaced 1 day apart (see how they are presented in Fig. 2. on the screen (5), 12 first representative images (100) spaced 24 hours apart, making these iterations, by selecting one of them, it is possible to recover second representative images (101) that are closer to it (see in Fig. 2 the presentation on the screen (5) of 12 second representative images (101)) in shorter time periods. , for example of 2 hours; and doing another iteration in the same way, that is, selecting one among the second representative images (101) recovered, third representative images (102) close to it will be recovered again, for example at intervals of 10 minutes (see also in fig. 2), with the limit established by the preset sending time of representative images from the cameras, which is preferably established precisely in a period of one minute between images, since in this way, At most, 1440 images per day will be sent to the server from each camera (less, if there is no motion detection), which greatly reduces storage needs as well as data transmission (video) through the communications networks used for this. Among any of the representative images (100, 101, 102) the marker image (A) can be selected. As indicated by way of preferred example, in these iteration substeps the search and retrieval of successive representative images close to the selected representative image can be carried out at times selected between 1 day, 2 hours, 10 minutes and 1 minute, for example. Once you select the image closest in time to the event you are looking for, you can proceed to download a short video fragment that includes said event.

For its part, motion detection can be done by:

- detection of scene variations in successive images of the video being recorded (significant differences in the recorded pixels) by a comparator of successive images associated with each camera, and/or

- motion detectors.

Notwithstanding the foregoing, and since the description made corresponds only to an example of a preferred embodiment of the invention, it will be understood that within its essentiality multiple variations of detail may be introduced, also protected, which may affect the shape, size or the manufacturing materials of the assembly or its parts, without this implying any alteration of the invention as a whole, limited: only by the claims provided below.

Claims (10)

1. System (1) for recording, locating and retrieving images for video surveillance, characterized in that it comprises: - a plurality of video surveillance cameras (2), - particular storage units (2a) in loop for the videos taken, associated with each of the cameras (2). - a motion detection module (2b) associated with each of the cameras (2). -an indexing server (3), provided with a memory (3a) for storing representative images (100, 101, 102) extracted from the videos taken by each camera (2), provided with a controller (4) for displaying the images images, associated with a display screen (5), - a management interface (4a), and - communications links (6) that are connecting each camera (2) and its particular storage unit (2a) with the indexing server (3). 2. System (1) for recording, locating and recovering images for video surveillance according to claim 1 where each motion detection module (2b) of the cameras (2) comprises a comparator of successive images. 3. System (1) for recording, locating and retrieving images for video surveillance according to claim 1 where each motion detection module (2b) comprises a motion detector. 4. Method of recording, locating and recovering images for video surveillance, characterized in that it includes - a stage of registration and indexing of representative images, which includes the following substages: - loop video recording from a plurality of cameras (2), of spaces intended to be monitored, and storing said videos in particular storage units (2a) of each camera (2), - detection of movements in the monitored spaces, - in case of positive movement detection, send representative images (100, 101, 102), extracted from the videos taken each adjusted sending time period, to an indexing server (3), through links ( 6) communications that are connecting each camera (2) and its particular storage unit (2a) with said indexing server (3), and - storage of the representative images (100, 101, 102) received on the indexing server (3) in a storage memory (3a); and an image search and display stage that includes the following substages: - introduction of a date (with time) or a range of dates for retrieval of representative images stored on the indexing server (3), through a controller (4) for displaying the images of the indexing server (3). - search on the indexing server (3) for representative images (100, 101, 102) close to the date and time entered, or to the range of dates and times entered, through the controller (4) for displaying the images. images, and presentation of said first representative images (100, 101, 102) recovered on a display screen (5). - manual selection, through the display controller (4), between the representative images (100, 101, 102) recovered from at least one marker image (A). and - download, from the particular storage unit (2a) of each camera (2) associated with the selected marker image (A), of a video fragment (A') that contains said marker image (A). 5. Method of recording, locating and recovering images for video surveillance according to claim 4, which comprises the search for first representative images (100), closest prior to the date and time entered, or prior to each date in a range of dates. introduced, and the presentation of said first representative images (100) recovered on the display screen (5). 6. Method of recording, locating and recovering images for video surveillance according to claim 5, where the search for the first representative images (100) comprises the recovery of the closest image prior to the date entered and a series of first representative images ( 100) following this at certain time intervals, and their presentation on the display screen (5) 7. Method of recording, locating and retrieving images for video surveillance according to any of claims 4 to 6, comprising one or more iteration sub-stages of searching for successive representative images (101, 102) increasingly closer in time to a previous representative image. 8. Method for recording, locating and recovering images for video surveillance according to claim 7, where in the iteration substages, the search and recovery of representative images close to the selected representative image is carried out, at times determined between 1 day, 2 hours, 10 minutes and 1 minute. 9. Method of recording, locating and retrieving images for video surveillance according to any of the previous claims, where the detection of movements is carried out by: - detection of scene variations in successive images of the video that is being recorded by a successive image comparator associated with each camera, - motion detectors 10. Method for recording, locating and retrieving images for video surveillance according to any of the preceding claims, where the time lapse of sending representative images to the indexing server (3) from each camera (2) is set to one minute.